[ HPF Home | Versions | Compilers | Projects | Publications | Applications | Benchmarks | Events | Contact ]

---

Simmux

Back to Applications

Will Dennisen,
TNO, denwja@tpd.tno.nl

Molecular Dynamics for simulating charged particles in fluid.

Number of Lines of Code: 5000

Target Platforms and HPF Compilers Used:

Platforms: CRAY, T3E, Pentium Clusters Compilers: pghpf, Cosy-hpf

Coding Styles (data decompositions, computational methods):

F77 code, structured, common blocks, irregular lists, load balancing.

Performance Information, if Available (including any possible comparisons to MPI and/or OpenMP):

Speed up, up to 4 processors (not fine tuned) with 1024 particles. 
Better performance for larger particle sizes.

Please comment on any aspects of the application that might be interesting, including any problems using HPF effectively:

Efficient implementation of global operations was a problem. Load balancing through adding ?processor dimension? to distributed arrays T(.., --, no proc)